Lucy Scott - Tokamak Energy
Ed Pinkney - Tokamak Energy
Fusion: An Introduction
Fusion is the reaction that powers the stars. At the centre of every star, two particles come together and fuse into one larger particle, which releases energy. In experiments on Earth, the nuclei of deuterium and tritium (types of heavy hydrogen) drive this process by fusing together to make a helium nucleus, a perfectly safe waste product. Fusion produces no greenhouse gases and no long-lived radioactive waste, yet it produces vast amounts of energy from very little fuel. As such it offers a clean, green, and plentiful energy solution if it can be harnessed on Earth, using the naturally present deuterium and tritium found in sea water.
For fusion reactions to occur, the deuterium and tritium must be very hot so particles can move at speeds high enough that they fuse upon collision. To safely create this fusion environment on Earth, we use specialised machines called tokamaks. Tokamaks use magnetic fields to keep the superheated fuel (now an electrically charged gas called plasma) from touching the walls of a vacuum chambers, keeping it hot enough for fusion to occur. If the plasma does hit the wall it is not dangerous, it simply cools down and stops the fusion process. Tokamak Energy aims to achieve these conditions in a spherical tokamak, which promises a smaller and cheaper solution than the larger tokamaks used in fusion experiments such as JET and ITER.
Since the possibility of controlled fusion was first demonstrated in the 1990s, the challenge to generate consumable energy for sustained periods using the fusion process has become an engineering challenge, rather than a physics problem. Tokamak Energy aims to accelerate the development of fusion energy by combining two emerging technologies: spherical tokamaks and high-temperature superconductors. Using this approach, we can design machines that are economical to build, run, and decommission.
We have broken down this huge challenge into the following five milestone stages:
- Build a small prototype tokamak to demonstrate the concept (achieved in 2013).
- Build a tokamak that uses magnetic coils constructed using a high-temperature superconductor (achieved in 2015).
- Reach fusion temperatures of 100 million degrees Celsius in a compact tokamak (predicted for 2018).
- Achieve energy breakeven conditions, in other words, generate amounts of energy equal to that required to drive the fusion process (predicted for 2020).
- Demonstrate a fusion power plant producing consumable electricity for the first time by 2025.
We are currently at Stage 3 and are building a new tokamak, the ST40, at our facility at Milton Park, Oxfordshire. We want to demonstrate that we can achieve fusion temperatures in a small tokamak. Stage 4 (energy breakeven conditions) will be a world-changing technology breakthrough.
The ST40 and Our Current Challenge
We are currently striving to build the ST40 from scratch. At this stage, we are building a tokamak that uses short bursts of the strong magnetic fields to create the conditions for fusion utilising super-capacitor power supplies and copper coils. We must complete this with only a few people, in a small building, and in a short time.
For the ST40 control system, we designed a scalable systems platform that integrates with our LabVIEW subsystems, CompactRIO systems, and third-party instrumentation. This gives us the option to increase functionality in the future. We use NI solutions throughout—from the implementation layer through to the server, data storage, and the control room screens. We want to oversee two control systems from the control room: the Machine Control System (MCS), and the Plasma Control System (PCS).
We are working to connect all our sub-systems to feed into a central control room. This will ensure we can efficiently operate and manage communications between all systems and manage data. It will also provide a platform for the control room equipment with many different screens and views.
Each time we drive the magnetics to create a fusion plasma is called a pulse. Before and during the pulse, the MCS monitors and controls almost all systems and displays them to the operations personnel using the control room screens. The MCS works between pulses to ensure the vacuum chamber is prepared and kept in a perfect state, ready for optimal plasma conditions.
- Running a pulse is a multistage process comprising the following steps, all operated from the control room:
- Super-capacitor based power supplies for the magnets are charged up ready for the electrical pulse. All the Power Supply Controllers connect to the LabVIEW MCS Server.
- The pulse ‘recipe’ is prepared, checked and then and loaded up into the PCS. The physicists set up the parameters to meet the objectives of the pulse.
- The MCS prepares all systems required for the pulse, cameras, gas supply, cooling, and diagnostics equipment.
Once everything is confirmed as ready, the MCS runs the pulse, managing all the peripheral systems and providing synchronised clocks to the diagnostics.
- The PCS is responsible for the fast feedback control of the fields during an experimental pulse. Using a mixture of feed-forward and feedback signals, the PCS tightly controls the current supplied to the coils that generate the magnetic fields in a tokamak vacuum chamber, so we can maintain plasma stability.
- Throughout the pulse, we measure the magnetic fields and plasma performance using many diagnostic sensors, including high-speed cameras, mass spectrometers, pick-up loops, temperature sensors, pressure sensors, and further advanced plasma measurements. Diagnostics racks collect several hundred diagnostics signals for post-analysis, using pre-existing third-party equipment. We can access this using a variety of techniques. We built a new library to access the data storage from LabVIEW to accommodate this.
- We collect large data sets during the experiment—one Gigabyte is generated per two-second pulse, with control room screens providing users with seamless interaction with this data.
Control System Design and Implementation
Our previous experience using LabVIEW was to control our previous Tokamak model (the ST25). We outline this in the case study titled “Using CompactRIO and LabVIEW to Monitor and Control a Compact Spherical Tokamak for Plasma Research.” From the start, we only considered NI technology. We have found that it delivers an incredibly scalable framework for building up a reliable control system that can communicate with a variety of controllers and sensors, as well as other more complicated or bespoke equipment. Our development team consists of two controls engineers (a Certified LabVIEW Architect and an experienced LabVIEW developer), working closely with a small physics and electrical team.
For our current work, we needed to measure all diagnostics with a synchronised clock. In this system, the plasma only gets hot for a few milliseconds and the complete pulse takes less than three seconds, so accurate and precise timing is critical. We currently use eight CompactRIO devices as our system controllers, with plans to include more in the future. With thousands of amps of current in our building, correct grounding is critical. Therefore, we placed the CompactRIO devices at various locations throughout our building.
Examples of our key systems include our system clock, the pulse control clock, and the vacuum vessel monitoring and control systems. The system clock and pulse control consist of two cRIO-9076 controllers. It uses several NI-9401 fast TTL modules connected to fibre optic converters for extremely accurate triggering. The vacuum vessel monitoring uses another CompactRIO controller with NI-9216 modules to simultaneously measure over 40 temperatures both inside the vessel and around the magnetic coils. This gives us excellent insight into conditions inside the vacuum chamber. Both during and outside of experimental runs, we use a system that includes a cRIO-9035 device to maintain chamber conditions through vacuum control, glow discharge cleaning and film deposition. This system operates firstly for long periods outside ‘pulses’ and secondly to control pre-ionisation during a pulse.
Meeting Our Targets: The Value of a Simplified Design
This project has aggressively short time frames and a small development team. We are currently preparing for our first full plasma trials and are confident in the solution we have chosen. By using LabVIEW and CompactRIO, we have designed a system that is scalable, reusable, and maintainable.
We have designed some special electronics ourselves, but there are many instances in which the ability to program the FPGA directly using a wide selection of standard I/O modules means that we can build high-speed solutions in hours, rather than weeks. By using standard components when possible, we simplified the solution. The flexibility of the CompactRIO platform will prove invaluable as we include new requirements in the project.
We have had help from NI support engineers, especially at the start of the project to review the original ST25 system and our proposed architecture. Because of this, we are pleased to be working with a reliable and responsive development environment as we commit to working at the cutting-edge of fusion research for a greener, sustainable future.
120A Olympic Avenue
Milton Park, Oxfordshire OX14 4SA, United Kingdom.